Ovarian cancer is the leading cause of death, among the gynecologic malignancies, and fifth leading cause of death, among cancers in women overall. These statistics largely reflect the fact that most women are diagnosed with ovarian cancer in advanced stages associated with poorer survival. With no certain method for primary prevention, early detection offers the best opportunity for immediately reducing mortality from this disease. However, pelvic ultrasound is costly, may lead to surgery for non-cancerous cysts, and may miss early lesions. Tumor markers, such as CA 125, adapted for screening also have uncertain sensitivity for early stage disease. In this application, investigators at the Brigham and Women's Hospital and Massachusetts General Hospital in Boston and St. Bartholomew's Hospital in London have and will further develop a variety of epidemiologic/specimen banks and pursue new approaches to the identification of serologic markers for ovarian cancer. The epidemiologic/specimen banks include: 1) epidemiologic data, plasma, buffy coat, and red cells from population-based case- control studies of women with newly diagnosed ovarian cancer; 2) epidemiologic data, pre-operative serum and plasma, and fresh frozen tissue on women with early stage ovarian cancer; 3) epidemiologic data, serum, and plasma from women who request genetic testing or who come to prophylatic oophorectomy through familial ovarian cancer clinics in Boston and London; and 4) epidemiologic data and sera from a population-based, longitudinal study designed to assess markers for the early detection of ovarian cancer. These data and specimens will be available both to Early Detection Research Network (EDRN) members as well as co- investigators within our project to pursue an integrated and novel approach to the early detection of ovarian cancer. Our approach will involve: 1) use of Surface-Enhanced Laser Desorption/Ionization (SELDI) mass spectrometry to screen for protein biomarkers that distinguish women with ovarian cancer from those without it or neoplastic from normal tissue; 2) development of statistical methods for combining multiple markers for ovarian cancer to identify panels of biomarkers with superior operating characteristics compared to any single biomarker; 3) purification, sequencing, and recombinant production of the most promising candidate panel of biomarkers to develop a high throughput ELISA or SELDI test kits; 4) assays of the most promising panel of biomarkers in the most informative specimens in the longitudinal screening database; and 5) generalization of statistical techniques to longitudinally measured data. Together with EDRN collaborators, our goal is to develop strategies for the early detection of ovarian cancer that are practical and efficient and can be projected to offer the best opportunity to reduce the mortality from a major cancer in women.